Ideas for really small slide switches?

On 17 Jan 2006 06:39:34 -0800, "larwe" <zwsdotcom@gmail.com> wrote:

>
>Paul Carpenter wrote:
>
>> Off the wall suggestion - forget mounting switches and use test pads
>> arranged as switches in SPDT configuration and a small metal stylus
>> to short to 1 or 0. I do assume you will have a switchscanner in
>> there in some form of other. This means building up switch patterns
>> and remembering each switch until an 'action' switch is operated.
>
>The scanning is no problem; as you can see I've already accepted the
>possibility of using momentary switches (pushbuttons) instead of "real"
>switches. This issue only affects the 35 switches below the IN-OUT
>lamps, by the way.
>
>The stylus is, however, a problem. If it comes to a point, it's hard to
>make it contact both sides of the "switch" contacts. If it comes to a
>flat surface, it needs to be aligned very well by the user so it
>presses hard enough against both contacts to make a good circuit.

You don't need 2 contacts. .
Use a single pad plus the user's body resistance to the part of the board that they are holding it
by. 

Roberto Waltman wrote:
> (Paul Carpenter) wrote:
> "larwe" wrote:
>
> That's very much on my wall, I was going to suggest
> something just one step above that. Stack together
> on an axis a few metallic plates with a profile
> resembling a switch, and use two pcb's to create
> a sensor pad on both sides.

I'm not sure I understand this. You're proposing contact plates on one
layer of the PCB, and this structure held into a routed hole in a
second layer of PCB material on top of it?

> You can have small parts made a-la-carte here:
>   http://www.emachineshop.com/

Very nifty site, although I don't care for the proprietary software.
Bookmarked!

On 17 Jan 2006 09:02:08 -0800, "larwe" <zwsdotcom@gmail.com> wrote:
>Roberto Waltman wrote:
>> (Paul Carpenter) wrote:
>> "larwe" wrote:
>>
>> That's very much on my wall, I was going to suggest
>> something just one step above that. Stack together
>> on an axis a few metallic plates with a profile
>> resembling a switch, and use two pcb's to create
>> a sensor pad on both sides.
>
>I'm not sure I understand this. You're proposing contact plates on one
>layer of the PCB, and this structure held into a routed hole in a
>second layer of PCB material on top of it?

+ or -  (Was in a hurry) I thought the
"switch assembly" could stand alone,
above the PCB, but your "routed hole
on a second pcb" will definitely make
it look better.

I thought two small pcb strips (not in my
drawing) held vertically on both sides along
the bottom of the "switches" would provide
the contacts. Small metal brackets soldered
to an horizontal PCB (if we think of the
switches as vertical) could do the same.

Sorry if this is not clear, I'll make a
drawing when I get some time on a system
with a CAD or ray-tracing program.

( It will take a while, this week I am
undergoing training on how to use an
unnecessarily complicated 3rd party
communication library in one of our
products. The kind that takes 3000+
lines to write a basic "Hello world"
program... )

Roberto Waltman

[ Please reply to the group, ]
[ return address is invalid. ]

Roberto Waltman wrote:

> Sorry if this is not clear, I'll make a
> drawing when I get some time on a system
> with a CAD or ray-tracing program.

Please don't take this much trouble on the matter - I am currently
looking at the Alps SKRK series 2.9mm x 3.9mm switches which might just
fit the application.

> undergoing training on how to use an
> unnecessarily complicated 3rd party
> communication library in one of our

I know the feeling only too well, and commiserate.

In article <20060117.1354.314415snz@pcserviceselectronics.co.uk>, paul
$@pcserviceselectronics.co.uk says...
> On 16 Jan, in article
>      <1137460695.985004.217880@g49g2000cwa.googlegroups.com>
>      zwsdotcom@gmail.com "larwe" wrote:
> 
> >I'm trying to build the smallest possible tangible PDP-1 emulator. My
> >ideal scenario would fit the circuit onto a double-sided 90 x 55mm PCB.
> >
> >Hence, I need to fit 119 lamps (LEDs) and 44 switches into this area.
> >The LEDs are no problem, but are there such things as tiny, tiny slide
> >switches or subminiature DIP switches? The smallest I can find are SMD
> >DIP switches with .050" lead spacing; is there anything smaller than
> >this?
> >
> >(It does not have to be practical. Merely amusing, and theoretically
> >operable).
> 
> Off the wall suggestion - forget mounting switches and use test pads
> arranged as switches in SPDT configuration and a small metal stylus
> to short to 1 or 0. I do assume you will have a switchscanner in
> there in some form of other. This means building up switch patterns
> and remembering each switch until an 'action' switch is operated.
> 
> Might save a lot of space.
> 
> 

How about an unplated through-hole with half of the hole one contact, the other half 
another contact? 

Huh, hard to describe. 

If you have a regular plated through hole on a PCB, then poking a meter probe at the hole 
for a voltage measurement (for instance) results in the tip of the lead centering itself 
in the hole. A stylus of the right shape would do the same thing. 

Now if half of that hole were one node and the other half another node, then the metal 
stylus would make contact between the two nodes and complete a circuit. The tendency for 
the stylus to center itself would help assure contact. 

The hole would have a half ring ( 170 degrees of a circle ) on one side and would be 
insulated from the half ring on the other. The stylus contact makes the connection. 0$ 
spent on switches.

Hello Lewin,


>                                           ... (NTSC is a
> dying standard - I don't want to design for it).
> 

Nah, all the digital cameras use it. Well, that and PAL. OTOH I can't 
imagine anyone crawling behind their PC, tipping over the coffee mug 
while doing that, wiggling the bolts and unplugging the VGA cable. Plus 
how do you connect that garden hose to your little card?


>>Hmm...  In the act of "pushing" the switches would it be acceptable to
>>consider finding a way of sensing if an LED is touched or pressed and
>>using the LED itself as sensor as well as display?
> 
> Interesting, I didn't consider this as a possibility (mind you I have
> no idea how it could be done). The idea of using pushbuttons is OK, I
> guess. I can have an LED by each to indicate if it is off or on.
> 

Maybe during brief non-lit phases. The uC would have to go hi-Z and then 
something would need to detect the charge change by an approaching finger.

How about a very thin metallization on top of each LED and using that as 
a touch sensor? Thin enough to see through but conductive enough to lead 
to the circuit board or LED pad and generate a large enough charge hump.

Regards, Joerg

http://www.analogconsultants.com

Joerg wrote:

> >                                           ... (NTSC is a
> > dying standard - I don't want to design for it).
> >
>
> Nah, all the digital cameras use it. Well, that and PAL. OTOH I can't

In a few years you won't be able to buy an NTSC television set. I don't
like designing for a standard that already has an execution date.

> imagine anyone crawling behind their PC, tipping over the coffee mug
> while doing that, wiggling the bolts and unplugging the VGA cable. Plus
> how do you connect that garden hose to your little card?

Same problem with NTSC, by the way, and same solution - the card will
have a flat connector of the type used on LCDs, sold by JAE and Hirose
among others - and an adapter cable provided separately. Note that I
don't mean an FFC connector; the cables used are more usually
individual strands of very fine wire.

As for the PC issue - I am assuming that anyone with the remotest
interest in this has spare monitors lying around all over the place :)
If I cared to collect them, I could have 20+ 15" and 17" CRT monitors
per week, that is AT MINIMUM how many I see thrown away at work and on
the streets I traverse while walking my dog.

> How about a very thin metallization on top of each LED and using that as

Practical to do by hand?!!

I could build a series of digital theremins underneath the board and
infer from phase changes where the finger is. However essentially this
algorithm (if I remember correctly) is patented by LeapFrog, they use
it in their "touch-sensitive" (without switches) books and globes.

On Wed, 18 Jan 2006 02:18:01 GMT, Joerg
<notthisjoergsch@removethispacbell.net> wrote:

><snip>
>How about a very thin metallization on top of each LED and using that as 
>a touch sensor? Thin enough to see through but conductive enough to lead 
>to the circuit board or LED pad and generate a large enough charge hump.

Actually, my first thought was to use an LED pen and to pulse the LEDs
on the board used as lights, so that they could be "turned around" and
used as detectors, intermittently between on-pulses.  Second thought
was to consider an unpowered pen with a bit of fluorescent material
taken from the inside of those compact fluorescent bulbs (cheap --
just break one) because the LED would energize the material and it
would emit a decaying pulse of return energy when brought close to a
pulsing LED and this extra duration could be observed.

Hmm....  I think I'll work on the idea ... hybrid LED switches!

Jon

Jonathan Kirwan wrote:
> Joerg <notthisjoergsch@removethispacbell.net> wrote:
> 
>> <snip> How about a very thin metallization on top of each LED
>> and using that as a touch sensor? Thin enough to see through
>> but conductive enough to lead to the circuit board or LED pad
>> and generate a large enough charge hump.
> 
> Actually, my first thought was to use an LED pen and to pulse
> the LEDs on the board used as lights, so that they could be
> "turned around" and used as detectors, intermittently between
> on-pulses.  Second thought was to consider an unpowered pen with
> a bit of fluorescent material taken from the inside of those
> compact fluorescent bulbs (cheap -- just break one) because the
> LED would energize the material and it would emit a decaying
> pulse of return energy when brought close to a pulsing LED and
> this extra duration could be observed.
> 
> Hmm....  I think I'll work on the idea ... hybrid LED switches!

Reverse it.  The leds will almost certainly be driven by a scanning 
matrix.  Sense the position with a light pen.  This handles all the 
on leds.  Maybe the offs can receive a much shorter pulse, and/or a 
much lower repetition rate.

-- 
"If you want to post a followup via groups.google.com, don't use
  the broken "Reply" link at the bottom of the article.  Click on
  "show options" at the top of the article, then click on the
  "Reply" at the bottom of the article headers." - Keith Thompson
More details at: <http://cfaj.freeshell.org/google/>

On Wed, 18 Jan 2006 08:58:21 -0500, "Chuck F. " <cbfalconer@yahoo.com>
wrote:

>Jonathan Kirwan wrote:
>> Joerg <notthisjoergsch@removethispacbell.net> wrote:
>> 
>>> <snip> How about a very thin metallization on top of each LED
>>> and using that as a touch sensor? Thin enough to see through
>>> but conductive enough to lead to the circuit board or LED pad
>>> and generate a large enough charge hump.
>> 
>> Actually, my first thought was to use an LED pen and to pulse
>> the LEDs on the board used as lights, so that they could be
>> "turned around" and used as detectors, intermittently between
>> on-pulses.  Second thought was to consider an unpowered pen with
>> a bit of fluorescent material taken from the inside of those
>> compact fluorescent bulbs (cheap -- just break one) because the
>> LED would energize the material and it would emit a decaying
>> pulse of return energy when brought close to a pulsing LED and
>> this extra duration could be observed.
>> 
>> Hmm....  I think I'll work on the idea ... hybrid LED switches!
>
>Reverse it.  The leds will almost certainly be driven by a scanning 
>matrix.  Sense the position with a light pen.  This handles all the 
>on leds.  Maybe the offs can receive a much shorter pulse, and/or a 
>much lower repetition rate.

Geeze, I don't see how Lewin can fail.  Ideas are flowing!  I need a
PDP-11 in that form factor and I know that Lewin won't get to the
PDP-11 before he gets that darned PDP-1 done, so.... anything to push
him along....

Jon